2. Stewardship functions and components: a framework for analysis

Designing effective health policies is a complex endeavour, as health systems are characterised by numerous market failures, competing stakeholder interests, and multidimensional performance objectives (Roberts, Reference Roberts, Hsiao, Berman and Reich2004). Policymakers must balance the often-conflicting goals of expanding access, controlling costs, and assuring quality. Given this complexity, policy failures frequently stem from design flaws such as unrealistic assumptions, insufficient evidence, lack of specificity, and poor targeting (Béland et al., Reference Béland, Rocco and Waddan2018; McConnell, Reference McConnell2010b), as well as ineffective implementation, including poor coordination, mistakes in the use of discretion, and inefficient regulation (Bayat et al., Reference Bayat, Kashkalani, Khodadost, Shokri, Fattahi, Seproo, Younesi and Khalilnezhad2023; Dong et al., Reference Dong, Christensen and Painter2014; Hudson et al., Reference Hudson, Hunter and Peckham2019).

Overcoming these design and implementation challenges requires robust government stewardship. The notion of stewardship positions the government as the ‘trustee’ of public health, emphasising its duty to act in the interest of the population rather than just its own bureaucratic imperatives (Saltman and Ferroussier-Davis, Reference Saltman and Ferroussier-Davis2000). Effective stewardship enables the government to establish a strategic direction for the health system, mobilise stakeholders, and formulate workable policies in the design process. It also facilitates the identification of various deficiencies that impede health reforms (Brinkerhoff et al., Reference Brinkerhoff, Cross, Sharma and Williamson2019).

However, effective health system stewardship is a complicated undertaking that requires intricate coordination of various functions spanning policy design and implementation. Veillard et al. (Reference Veillard, Cowling, Bitton, Ratcliffe, Kimball, Barkley, Mercereau, Wong, Taylor, Hirschhorn and Wang2017) and Brinkerhoff (Reference Brinkerhoff, Cross, Sharma and Williamson2019) provide foundational works that illuminate the core aspects of stewardship in the healthcare sector. Building upon their insights and the work of Cross et al. (Reference Cross, Sayedi, Irani, Archer, Sears and Sharma2017, Reference Cross, De La Cruz and Dent2019), this paper proposes six essential stewardship functions that are instrumental in realising the transformative potential of health reform. These essential functions include (1) formulating strategic policy direction, (2) ensuring the alignment of system structures with strategy and policy goals, (3) establishing the legal, regulatory, and policy instruments to guide performance, (4) building and sustaining relationships, coalitions, and partnerships, (5) ensuring accountability and transparency, and (6) generating intelligence. To successfully carry out these functions, governments must cultivate and employ capacities to carry out a mix of tasks across three core components: political, analytical, and operational.

Table 1 presents a framework that integrates the six essential stewardship functions to their respective political, operational, and analytical requirements. This framework serves as a tool for assessing the effectiveness of stewardship in the context of health reform. By systematically examining each function and its associated components, policymakers and researchers can identify strengths, weaknesses, and opportunities for improvement in the stewardship of health systems.

Table 1. Stewardship functions and the requisite political, analytical, and operational components

Note: authors, based on Cross et al. (Reference Cross, Sayedi, Irani, Archer, Sears and Sharma2017, Reference Cross, De La Cruz and Dent2019).

The first function involves setting a strategic policy direction for the health system. This requires establishing a clear, compelling vision for reform that articulates the rationale, goals, and pathways for change (Reich, Reference Reich2020). Crafting a purposeful reform agenda demands robust situational analysis to diagnose health system shortcomings, identify windows of opportunity, and assess political feasibility (Roberts et al., Reference Roberts, Hsiao, Berman and Reich2004). Equally important is the ability to mobilise diverse stakeholders around a shared reform narrative.

The second key function is aligning system structures with reform objectives. Health systems are inherently complex, with multiple actors with divergent interests operating across different levels of governance (Frenk et al., Reference Frenk, Chen, Bhutta, Cohen, Crisp, Evans, Fineberg, Garcia, Ke, Kelley, Kistnasamy, Meleis, Naylor, Pablos-Mendez, Reddy, Scrimshaw, Sepulveda, Serwadda and Zurayk2010). Implementing reforms requires reconfiguring institutional arrangements and organisational incentives to orient behaviour toward desired outcomes. Structural alignment may involve changes to authority hierarchies, service delivery models, financing flows, and accountability relationships (Figueras, Reference Figueras, Robinson and Jakubowski2005).

Selecting appropriate policy instruments to guide health system performance is the third critical stewardship function. Policy instruments refer to the specific rules, standards, incentives and nudges that shape actor behaviour within a given health system (McDonnell and Elmore, Reference McDonnell and Elmore1987). Designing effective instruments requires a keen understanding of how different policy levers interact with each other and with local contexts to produce intended and unintended consequences. This complexity also makes it imperative to complement policy design with robust, evidence-based implementation strategies that can navigate competing priorities while operating within resource constraints.

The fourth function, stewarding health reforms, demands active coalition-building and partnership management. Engaging diverse stakeholders – including providers, payers, patients, and civil society organisations – is essential for mobilising resources, fostering collective action, and generating reform legitimacy (Rasanathan et al., Reference Rasanathan, Atkins, Mwansambo, Soucat and Bennett2018). The process begins with a careful analysis of the stakeholder landscape, power dynamics and potential points of conflict, which informs the development of collaboration frameworks. Nurturing such multi-sectoral collaborations can enhance policy coherence, implementation coordination, and mutual accountability (Bennett et al., Reference Bennett, Glandon and Rasanathan2018).

The fifth key function involves institutionalising accountability and transparency mechanisms. Accountability refers to the processes by which health system actors are held responsible for their actions and outcomes (Brinkerhoff, Reference Brinkerhoff2004). Effective accountability systems combine top-down oversight tools with bottom-up feedback channels and operate across three broad areas: financial, system performance, and political (Cross et al., Reference Cross, Sayedi, Irani, Archer, Sears and Sharma2017; Joshi, Reference Joshi2013). Transparency, applied both within health ministries and in their public interactions, entails the proactive disclosure of health system information to enable improvement opportunities and public scrutiny and promote trust (Paschke et al., Reference Paschke, Dimancesco, Vian, Kohler and Forte2018).

Finally, stewarding health reforms requires continuous intelligence-gathering to inform adaptation. Generating timely, relevant data on health system performance is crucial for surfacing implementation challenges, identifying course corrections, and demonstrating progress (Wyber et al., Reference Wyber, Vaillancourt, Perry, Mannava, Folaranmi and Celi2015). This may include investing in health information systems, commissioning targeted evaluations, and institutionalising learning feedback loops. Equally important is how this evidence is translated into action – informing both policy decisions and public dialogue to drive meaningful change.

While the specific manifestations of stewardship may vary across contexts, the core functions identified in this framework serve as a generalisable template for assessing and strengthening reform processes. The following section outlines the methods employed in this study before applying the analytical framework to examine how stewardship operated in practice during China’s Sanming health reform. By tracing stewardship activities across each functional domain, the case analysis generates transferable lessons for reform practitioners grappling with similar challenges in other settings.

3. Research methods

This study employs a qualitative case study approach to examine how active stewardship shapes health reform outcomes, using Sanming as an illustrative example. Sanming, a prefecture-level city in China’s Fujian province, provides an ideal case for analysis given its significant reform achievements in curbing cost growth, strengthening primary care, and improving public hospital performance.

The investigation is guided by an analytical framework derived from health systems and public policy literature, which identifies six core stewardship functions: strategic visioning, institutional alignment, instrument design, partnership management, accountability reinforcement, and learning facilitation. This framework structures the analysis of how specific stewardship practices influenced both the design and implementation of Sanming’s reforms.

To ensure methodological rigour, we employed a triangulated research approach combining document analysis with primary data collection. The document analysis encompassed a comprehensive review of published research, policy documents, and government reports concerning Sanming’s reforms. We systematically searched PubMed, SCOPUS, and Web of Science databases, alongside grey literature repositories, using keywords including ‘Sanming’, ‘health reform’, ‘health policy’, ‘stewardship’, and ‘China’. Both English and Chinese materials analysing governance factors within the Sanming reform context were included, providing detailed insights relevant to each of the six framework domains.

Primary data collection involved semi-structured interviews conducted in April 2024. With local government facilitation, we interviewed 26 stakeholders directly involved in Sanming’s healthcare reform. This sample comprised six senior health policy officials from Sanming’s government, fourteen hospital administrators and doctors from three county hospitals, and six healthcare professionals from township and village facilities (Table 2). Purposive sampling ensured representation across key stakeholder groups, including policymakers from departments responsible for finance and public health policy formulation, and healthcare providers from the three-tier medical conglomerates through which reforms were implemented. To accurately capture the evolution of the Sanming reform, our methodology engaged key participants who had demonstrated exceptional continuity despite the reform spanning a decade. The core government policymakers have provided continuous leadership since the reform’s inception, while senior hospital administrators and doctors have implemented the reforms throughout this period. This approach ensures that our data provides a consistent, longitudinal, and authoritative perspective on the evolution of the reform.

Table 2. Interview respondents

Note: Hospitals B and E, along with primary care facility G, belong to the same medical conglomerate; Hospitals C and F belong to the same medical conglomerate.

The semi-structured interview format balanced consistency with flexibility. We provided core topics while encouraging participants to explore issues they deemed significant through follow-up questions. Government officials discussed their decision-making processes, implementation challenges, and policy considerations. Healthcare providers at county, township, and village levels shared their interpretations of reforms, described practical implementation experiences, and reflected on operational difficulties. Interviews ranged from 20 to 60 minutes and, with participant consent, were recorded and transcribed verbatim.

Data analysis followed an iterative process of extraction, synthesis, and interpretation. Interview transcripts were analysed within the six-function stewardship framework to understand how these functions interact to shape reform success. Information on reform context, objectives, strategies, and outcomes was systematically organised according to each stewardship function. Through comparative analysis, key patterns and themes emerged, with illustrative examples selected to demonstrate practical application of stewardship functions. The analysis also examined interactions and tensions between different stewardship practices and their evolution over time. Interview excerpts were systematically integrated to substantiate key findings.

The following section reviews Sanming’s health reforms and analyses its experience through the developed framework, focusing on the core components of stewardship functions. The concluding section then discusses the active stewardship required for successful health reform.

4. Background and extant knowledge

Market-oriented reforms in China during the 1990s led to the dismantling of the rural cooperative medical system which in turn caused rising out-of-pocket costs and widening health inequities. In response, China launched a series of reforms in the early 2000s aimed at expanding insurance coverage, strengthening primary care, and improving service delivery (Ramesh and Bali, Reference Ramesh and Bali2021). Reforms in 2009 further expanded insurance coverage, increased government subsidies, and introduced a national essential medicines system. Despite progress, challenges related to quality, efficiency, and integration persisted. In recent years, China has scaled up the reforms by adopting broader and deeper reforms aimed at promoting value-based care, strengthening provider payment systems, and leveraging digital technologies. In Sanming, similar to many other cities, the reforms were distinguished by extensive piloting before being scaled up and expanded. What distinguished Sanming was the notable success its reforms centred on active government stewardship achieved, making it a fruitful case for study and emulation (He, Reference He2023).

Beginning in 2011, the Sanming municipal government embarked on an ambitious reform agenda aimed at controlling cost growth, improving service delivery, and improving the performance of public hospitals. The immediate cause of the reform was the bankruptcy of its medical insurance fund, with the medical insurance schemes together owed the public hospitals 17.5 million yuan (He, Reference He2018). Sanming has a seriously ageing population, with a large number of retired workers. This puts significant strain on the financial sustainability of the Urban Employee Basic Medical Insurance (UEBMI) and Urban and Rural Resident Basic Medical Insurance (URRBMI). In 2023, Sanming’s UEBMI dependency ratio was 1.47:1, meaning 1.47 active workers supported each retiree who had not contributed to the UEBMI system. This was far below the national average of 2.71:1, and the ratio continued to deteriorate, dropping to 1.40:1 in 2024. Furthermore, Sanming city lags behind in terms of economic development, ranking seventh out of the nine prefecture-level cities in Fujian Province. Consequently, the financing level of its funds remains relatively low, with per capita funding for medical insurance schemes for employees and residents falling below the national average. UEBMI stands at 4,204 yuan and URRBMI at 990 yuan per year, compared to national figures of 6,182 yuan and 1,098 yuan, respectively.

Despite these limitations, Sanming’s reforms have achieved notable success in ensuring the sustainability of the insurance fund and reducing out-of-pocket costs for patients. Since 2012, the UEBMI and URRBMI funds in Sanming have remained in surplus, while inpatient reimbursement rates have increased substantially, from 72.26 per cent to 76.63 per cent for urban employees and from 46.25 per cent to 68.63 per cent for residents, between 2011 and 2023 (Sanming Municipal Health Commission, 2024). Significant progress has also been made in chronic disease management, with control rates for hypertension and diabetes reaching 88.04 per cent and 87.54 per cent, respectively. Furthermore, provider prescription behaviours also shifted, with a notable decline in antibiotic prescription rate in county-level hospitals (Hu et al., Reference Hu, Fu, Wushouer, Ni, Li, Guan and Shi2022).

These achievements affirmed the local success of the reform and captured the attention of the central government, which played a crucial role in scaling up the Sanming pilot innovation. The Ministry of Finance initiated this state-led diffusion process when it identified the model’s potential to alleviate systemic financial strain. Subsequently, the National Health Commission encouraged the nationwide adoption of Sanming’s experience within five years to magnify the impacts of its pilots. (Song and Li, Reference Song and Li2024a; The State Council Information Office of the People’s Republic of China, 2024). To support this initiative, the central government has developed an integrated set of policy instruments spanning four key areas: governance, drug procurement, medical insurance, and service provider (Song and Li, Reference Song and Li2024b). This coordinated, top-down promotion has led to widespread, though uneven, adoption across China. Many municipalities have implemented selected components of the reform, particularly those with politically ambitious leadership or those experiencing fiscal pressures similar to Sanming’s initial situation. However, some local governments have adopted a cautious wait-and-see approach, observing the outcomes in regions that adopted the reform early before committing to full implementation (Li and Song, Reference Li and Song2023).

Overall, Sanming’s experience demonstrate that systemic reform can yield substantial gains even under constrained fiscal and demographic conditions. While the specific contours of Sanming’s reform reflect the idiosyncrasies of the local context, the underlying stewardship practices offer transferable lessons for policymakers in other settings, particularly in middle – and low-income countries. In the discussion below, we examine Sanming’s reforms in relation to six core stewardship functions to reveal how these stewardship practices contributed to significant improvements in health system performance (Wang et al., Reference Wang, Mao, Li and Du2015; Zhan, Reference Zhan2021).

5. The Sanming healthcare reform: active stewardship in practice

6. Conclusion: active stewardship for health reform success

Health systems worldwide face mounting pressures from epidemiological, technological, and sociopolitical shifts, making effective reform increasingly urgent. This paper demonstrates that active government stewardship across six core functions – strategic visioning, institutional realignment, instrumental calibration, coalition-building, accountability reinforcement, and adaptive learning – critically determines healthcare reform success. The Sanming experience illustrates how these functions interact to transform a fragmented, inefficient system into one that delivers higher quality, equitable, and affordable care.

Sanming’s reform succeeded through integrated active stewardship. A clear strategic vision aligned with local priorities was operationalised through institutional restructuring that enabled rather than constrained change. Calibrated policy instruments – payment reforms, incentive mechanisms, and service delivery models – provided behavioural change levers, while coalition-building across levels and sectors created alliances with both motivation and influence to sustain momentum. Accountability systems combining top-down oversight with bottom-up transparency ensured alignment, while continuous learning enabled adaptive refinement. These components formed a self-reinforcing cycle: strategic direction shaped institutional design, facilitating policy development and implementation, strengthened through stakeholder engagement, maintained through accountability, and refined through ongoing learning.

Sanming operates within China’s distinct political-institutional context, where centralised hierarchy may have facilitated decisive reform leadership. More pluralistic settings with diffused power would likely face greater implementation challenges. Yet China’s bureaucracy itself presents obstacles – its fragmented architecture creates coordination challenges (He, Reference He2018). Sanming addressed this by establishing a leading group consolidating authority across healthcare departments, demonstrating how active stewardship overcomes structural diffusion.

Framework transferability depends on adapting core functions to different governance architectures. Federal systems with distributed health governance require enhanced institutional alignment through intergovernmental agreements or joint committees. Mixed public-private systems can adapt strategic purchasing and performance accountability through regulatory rather than direct control mechanisms. The phased approach – cost containment, provider incentive realignment, then integrated delivery systems – offers a sequenced pathway calibrated to varying starting conditions.

The findings significantly impact health systems struggling with market-led or network-based governance limitations. Active government stewardship provides a viable alternative where fragmentation and marketisation undermine equity and efficiency, particularly relevant for decentralised, market-driven systems like India’s. China’s strengthened hierarchical governance through Sanming yielded tangible improvements in cost control and performance (Ramesh et al., Reference Ramesh, Wu and He2014; Ramesh et al., Reference Ramesh, Wu and Howlett2015).

The framework can be applied to diverse typologies. Middle-income countries pursuing universal coverage learn that resource constraints need not preclude ambitious reform when stewardship functions align properly. Systems with expanded insurance but persistent quality and sustainability challenges benefit from understanding how procurement reforms, provider realignment, and integrated delivery work synergistically. Intelligence generation through continuous monitoring particularly aids systems with robust information infrastructure but underutilised adaptive policymaking capacity.

Government involvement alone is insufficient, as sustainable reform requires integrating all six functions. This especially applies where universal coverage guarantees remain hampered by misaligned incentives and fragmented governance, resulting in uncoordinated implementation and inequitable quality, as exemplified by Brazil’s Unified Health System (Paim et al., Reference Paim, Travassos, Almeida, Bahia and Macinko2011; Coelho and Szabzon, Reference Coelho and Szabzon2024). Sanming’s coherent function alignment created reform-enhancing synergies.

Mature systems facing sustainability challenges can adapt Sanming’s approach. The medical conglomerate model shows how vertical integration addresses care fragmentation in coordinating primary-secondary care. Balancing centralised procurement with performance incentives informs pharmaceutical expenditure control while maintaining provider autonomy. Success requires thoughtful adaptation to existing architectures rather than wholesale adoption.

The framework transcends specific contexts by identifying fundamental governance capabilities. Whether in unitary or federal states, tax-funded or insurance-based systems, public or mixed delivery models, the six functions provide diagnostic tools for identifying gaps and designing interventions. This coordinated governance approach extends to education, social protection, and other complex domains requiring multi-level coordination.

This study has several limitations that should be acknowledged. The sampling strategy, which focused on government officials, hospital administrators, and medical professionals at county, township, and village levels, may have introduced self-selection bias. Moreover, the perspectives of external stakeholders – including patients, pharmaceutical firms, and private insurers – were not incorporated, potentially limiting the comprehensiveness of our findings. While triangulation with policy documents and official reports helped enhance validity and mitigate these biases, future research would benefit from incorporating these broader stakeholder perspectives. Comparative studies examining how different institutional contexts adapt stewardship functions would be particularly valuable, helping to distinguish which reform elements are universally transferable from those that remain context-specific.